‘Gambling with our lives’: Anthropic researcher quits cites fears of AI firms racing to self-improving superintelligence

According to Coxon the systems could soon have superhuman strength and this poses a level of danger to everyone.